Do Inpatient Antimicrobial Stewardship Programs Help Us in the Battle Against Antimicrobial Resistance?

Abstract

Background: Antibiotic stewardship programs (ASPs) have demonstrated success at reducing costs, yet there is limited quality evidence of their effectiveness in reducing infections of high-profile drug-resistant organisms.

Methods: This retrospective, cohort study included all Kaiser Permanente Southern California (KPSC) members aged ≥18 years hospitalized in 9 KPSC hospitals from 1 January 2008 to 31 December 2016. We measured the impact of staggered ASP implementation on consumption of 18 ASP-targeted antibiotics using generalized linear mixed-effects models. We used multivariable generalized linear mixed-effects models to estimate the adjusted effect of an ASP on rates of infection with drug-resistant organisms. Analyses were adjusted for confounding by time, cluster effects, and patient- and hospital-level characteristics.

Results: We included 765 111 hospitalizations (288 257 pre-ASP, 476 854 post-ASP). By defined daily dose, we found a 6.1% (-7.5% to -4.7%) overall decrease antibiotic use post-ASP; by days of therapy, we detected a 4.3% (-5.4% to -3.1%) decrease in overall use of antibiotics. The number of prescriptions increased post-ASP (1.04 [1.03-1.05]). In adjusted analyses, we detected an overall increase in vancomycin-resistant enterococci infections post-ASP (1.37 [1.10-1.69]). We did not detect a change in the rates of extended-spectrum beta-lactamase, carbapenem-resistant Enterobacteriaceae, and multidrug-resistant Pseudomonas aeruginosa infections post-ASP.

Conclusions: ASPs with successful reductions in consumption of targeted antibiotics may not see changes in infection rates with antibiotic-resistant organisms in the 2 to 6 years post-implementation. There are likely differing timescales for reversion to susceptibility across organisms and antibiotics, and unintended consequences from compensatory prescribing may occur.

Keywords: antibiotic consumption; antibiotic resistance; antibiotic stewardship.

Figure 1.

Adjusted percentage change in the geometric mean of (A) defined daily dose and (B) days of therapy. C, Adjusted relative risk of antibiotic use (yes/no) from pre-ASP to post-ASP. All models adjusted for period, medical center, age, race, sex, prior hospitalization, prior outpatient utilization, prior emergency department utilization, diagnosis-related group, Charlson group, infection on admission, and unit type. Anti-PSA antibiotics: imipenem, meropenem, amikacin, gentamicin, ceftazidime, cefepime, ciprofloxacin O, ciprofloxacin P, levofloxacin, aztreonam, piperacillin-tazobactam. Anti-MRSA antibiotics: daptomycin, linezolid, vancomycin. Community-onset antibiotics: ceftotaxime, ceftriaxone, ertapenem, moxifloxacin. Abbreviations: ASP, antibiotic stewardship program; CI, confidence interval; LCL, lower confidence limit; MRSA, methicillin-resistant Staphylococcus aureus; PSA, pseudomonal; RR, risk ratio; UCL, upper confidence limit.

Figure 1.

Adjusted percentage change in the geometric mean of (A) defined daily dose and (B) days of therapy. C, Adjusted relative risk of antibiotic use (yes/no) from pre-ASP to post-ASP. All models adjusted for period, medical center, age, race, sex, prior hospitalization, prior outpatient utilization, prior emergency department utilization, diagnosis-related group, Charlson group, infection on admission, and unit type. Anti-PSA antibiotics: imipenem, meropenem, amikacin, gentamicin, ceftazidime, cefepime, ciprofloxacin O, ciprofloxacin P, levofloxacin, aztreonam, piperacillin-tazobactam. Anti-MRSA antibiotics: daptomycin, linezolid, vancomycin. Community-onset antibiotics: ceftotaxime, ceftriaxone, ertapenem, moxifloxacin. Abbreviations: ASP, antibiotic stewardship program; CI, confidence interval; LCL, lower confidence limit; MRSA, methicillin-resistant Staphylococcus aureus; PSA, pseudomonal; RR, risk ratio; UCL, upper confidence limit.